Search

Summary: Rapid advancements in medical AI necessitate targeted educational initiatives for clinicians to ensure AI tools are safe and used effectively to improve patient outcomes. To support decision-making among stakeholders in medical education, we propose three tiers of medical AI expertise and outline the challenges for medical education at different educational stages. Additionally, we offer recommendations and examples, encouraging stakeholders to adapt and shape curricula for their specific healthcare setting using this framework.

Abstract Existing methods to characterise the evolving condition of traumatic brain injury (TBI) patients in the intensive care unit (ICU) do not capture the context necessary for individualising treatment. Here, we integrate all heterogenous data stored in medical records (1166 pre-ICU and ICU variables) to model the individualised contribution of clinical course to 6-month functional outcome on the Glasgow Outcome Scale -Extended (GOSE). On a prospective cohort (n = 1550, 65 centres) of TBI patients, we train recurrent neural network models to map a token-embedded time series representation of all variables (including missing values) to an ordinal GOSE prognosis every 2 h. The full range of variables explains up to 52% (95% CI: 50–54%) of the ordinal variance in functional outcome. Up to 91% (95% CI: 90–91%) of this explanation is derived from pre-ICU and admission information (i.e., static variables). Information collected in the ICU (i.e., dynamic variables) increases explanation (by up to 5% [95% CI: 4–6%]), though not enough to counter poorer overall performance in longer-stay (>5.75 days) patients. Highest-contributing variables include physician-based prognoses, CT features, and markers of neurological function. Whilst static information currently accounts for the majority of functional outcome explanation after TBI, data-driven analysis highlights investigative avenues to improve the dynamic characterisation of longer-stay patients. Moreover, our modelling strategy proves useful for converting large patient records into interpretable time series with missing data integration and minimal processing.

Abstract Background There is insufficient evidence to guide ventilatory targets in acute brain injury (ABI). Recent studies have shown associations between mechanical power (MP) and mortality in critical care populations. We aimed to describe MP in ventilated patients with ABI, and evaluate associations between MP and clinical outcomes. Methods In this preplanned, secondary analysis of a prospective, multi-center, observational cohort study (ENIO, NCT03400904), we included adult patients with ABI (Glasgow Coma Scale ≤ 12 before intubation) who required mechanical ventilation (MV) ≥ 24 h. Using multivariable log binomial regressions, we separately assessed associations between MP on hospital day (HD)1, HD3, HD7 and clinical outcomes: hospital mortality, need for reintubation, tracheostomy placement, and development of acute respiratory distress syndrome (ARDS). Results We included 1217 patients (mean age 51.2 years [SD 18.1], 66% male, mean body mass index [BMI] 26.3 [SD 5.18]) hospitalized at 62 intensive care units in 18 countries. Hospital mortality was 11% (n = 139), 44% (n = 536) were extubated by HD7 of which 20% (107/536) required reintubation, 28% (n = 340) underwent tracheostomy placement, and 9% (n = 114) developed ARDS. The median MP on HD1, HD3, and HD7 was 11.9 J/min [IQR 9.2–15.1], 13 J/min [IQR 10–17], and 14 J/min [IQR 11–20], respectively. MP was overall higher in patients with ARDS, especially those with higher ARDS severity. After controlling for same-day pressure of arterial oxygen/fraction of inspired oxygen (P/F ratio), BMI, and neurological severity, MP at HD1, HD3, and HD7 was independently associated with hospital mortality, reintubation and tracheostomy placement. The adjusted relative risk (aRR) was greater at higher MP, and strongest for: mortality on HD1 (compared to the HD1 median MP 11.9 J/min, aRR at 17 J/min was 1.22, 95% CI 1.14–1.30) and HD3 (1.38, 95% CI 1.23–1.53), reintubation on HD1 (1.64; 95% CI 1.57–1.72), and tracheostomy on HD7 (1.53; 95%CI 1.18–1.99). MP was associated with the development of moderate-severe ARDS on HD1 (2.07; 95% CI 1.56–2.78) and HD3 (1.76; 95% CI 1.41–2.22). Conclusions Exposure to high MP during the first week of MV is associated with poor clinical outcomes in ABI, independent of P/F ratio and neurological severity. Potential benefits of optimizing ventilator settings to limit MP warrant further investigation.

Disorders of consciousness are complex conditions characterised by persistent loss of responsiveness due to brain injury. They present diagnostic challenges and limited options for treatment, and highlight the urgent need for a more thorough understanding of how human consciousness arises from coordinated neural activity. The increasing availability of multimodal neuroimaging data has given rise to a wide range of clinically- and scientifically-motivated modelling efforts, seeking to improve data-driven stratification of patients, to identify causal mechanisms for patient pathophysiology and loss of consciousness more broadly, and to develop simulations as a means of testing in silico potential treatment avenues to restore consciousness. As a dedicated Working Group of clinicians and neuroscientists of the international Curing Coma Campaign, here we provide our framework and vision to understand the diverse statistical and generative computational modelling approaches that are being employed in this fast-growing field. We identify the gaps that exist between the current state-of-the-art in statistical and biophysical computational modelling in human neuroscience, and the aspirational goal of a mature field of modelling disorders of consciousness; which might drive improved treatments and outcomes in the clinic. Finally, we make several recommendations for how the field as a whole can work together to address these challenges.

Abstract Purpose External ventricular drains (EVD) are commonly used in aneurysmal subarachnoid hemorrhage (aSAH) patients and can be life-saving by diverting cerebrospinal fluid. However, the overall relationship between EVD use and outcome is poorly understood. Methods In an exploratory analysis of an aSAH patient cohort, we examined EVD use in relation to modified Rankin Scale (mRS) at hospital discharge and at 6 months (unfavorable outcome = mRS > 2) using univariable and multivariable analyses. Results EVDs were placed in 31 of 56 (55.4%) patients and more often in women than men (66.7% vs 35.0%, p = 0.022) despite similar rates of hydrocephalus. Women had greater ICU [18 (13.5–25) vs 11.5 (6.5–18.5) days, p = 0.014] and hospital lengths of stay (LOS) [20.5 (16.5–34) vs 13.5 (10.5–27) days, p = 0.015] than men and greater mRS at discharge [4 (3–5) vs 3 (2–3.5), p = 0.011] although mRS at 6 months was similar. Patients with EVDs had longer ICU and hospital LOS and greater mRS at discharge [5 (3–6) vs 2 (2–3), p

Statin therapy is associated with numerous cellular effects. Here, the authors show that statin treatment increases post-translational modifications on fatty acid synthase in the active site, revealing communication between the cholesterol and lipid biosynthetic pathways.

Abstract Our goal is to explore quantitative motor features in critically ill patients with severe brain injury (SBI). We hypothesized that computational decoding of these features would yield information on underlying neurological states and outcomes. Using wearable microsensors placed on all extremities, we recorded a median 24.1 (IQR: 22.8–25.1) hours of high-frequency accelerometry data per patient from a prospective cohort (n = 69) admitted to the ICU with SBI. Models were trained using time-, frequency-, and wavelet-domain features and levels of responsiveness and outcome as labels. The two primary tasks were detection of levels of responsiveness, assessed by motor sub-score of the Glasgow Coma Scale (GCSm), and prediction of functional outcome at discharge, measured with the Glasgow Outcome Scale–Extended (GOSE). Detection models achieved significant (AUC: 0.70 [95% CI: 0.53–0.85]) and consistent (observation windows: 12 min–9 h) discrimination of SBI patients capable of purposeful movement (GCSm > 4). Prediction models accurately discriminated patients of upper moderate disability or better (GOSE > 5) with 2–6 h of observation (AUC: 0.82 [95% CI: 0.75–0.90]). Results suggest that time series analysis of motor activity yields clinically relevant insights on underlying functional states and short-term outcomes in patients with SBI.

Abstract Traumatic brain injury (TBI) is a leading neurological cause of death and disability across the world. Early characterization of TBI severity could provide a window for therapeutic intervention and contribute to improved outcome. We hypothesized that granular electronic health record data available in the first 24 h following admission to the intensive care unit (ICU) can be used to differentiate outcomes at discharge. Working from two ICU datasets we focused on patients with a primary admission diagnosis of TBI whose length of stay in ICU was ≥ 24 h (N = 1689 and 127). Features derived from clinical, laboratory, medication, and physiological time series data in the first 24 h after ICU admission were used to train elastic-net regularized Generalized Linear Models for the prediction of mortality and neurological function at ICU discharge. Model discrimination, determined by area under the receiver operating characteristic curve (AUC) analysis, was 0.903 and 0.874 for mortality and neurological function, respectively. Model performance was successfully validated in an external dataset (AUC 0.958 and 0.878 for mortality and neurological function, respectively). These results demonstrate that computational analysis of data routinely collected in the first 24 h after admission accurately and reliably predict discharge outcomes in ICU stratum TBI patients.

Abstract Background Cell transplantation-based treatments for neurological disease are promising, yet graft rejection remains a major barrier to successful regenerative therapies. Our group and others have shown that long-lasting tolerance of transplanted stem cells can be achieved in the brain with systemic application of monoclonal antibodies blocking co-stimulation signaling. However, it is unknown if subsequent injury and the blood-brain barrier breach could expose the transplanted cells to systemic immune system spurring fulminant rejection and fatal encephalitis. Therefore, we investigated whether delayed traumatic brain injury (TBI) could trigger graft rejection. Methods Glial-restricted precursor cells (GRPs) were intracerebroventricularly transplanted in immunocompetent neonatal mice and co-stimulation blockade (CoB) was applied 0, 2, 4, and 6 days post-grafting. Bioluminescence imaging (BLI) was performed to monitor the grafted cell survival. Mice were subjected to TBI 12 weeks post-transplantation. MRI and open-field test were performed to assess the brain damage and behavioral change, respectively. The animals were decapitated at week 16 post-transplantation, and the brains were harvested. The survival and distribution of grafted cells were verified from brain sections. Hematoxylin and eosin staining (HE) was performed to observe TBI-induced brain legion, and neuroinflammation was evaluated immunohistochemically. Results BLI showed that grafted GRPs were rejected within 4 weeks after transplantation without CoB, while CoB administration resulted in long-term survival of allografts. BLI signal had a steep rise following TBI and subsequently declined but remained higher than the preinjury level. Open-field test showed TBI-induced anxiety for all animals but neither CoB nor GRP transplantation intensified the symptom. HE and MRI demonstrated a reduction in TBI-induced lesion volume in GRP-transplanted mice compared with non-transplanted mice. Brain sections further validated the survival of grafted GRPs and showed more GRPs surrounding the injured tissue. Furthermore, the brains of post-TBI shiverer mice had increased activation of microglia and astrocytes compared to post-TBI wildtype mice, but infiltration of CD45+ leukocytes remained low. Conclusions CoB induces sustained immunological tolerance towards allografted cerebral GRPs which is not disrupted following TBI, and unexpectedly TBI may enhance GRPs engraftment and contribute to post-injury brain tissue repair.

Abstract Coronavirus SARS-COV-2 infections continue to spread across the world, yet effective large-scale disease detection and prediction remain limited. COVID Control: A Johns Hopkins University Study, is a novel syndromic surveillance approach, which collects body temperature and COVID-like illness (CLI) symptoms across the US using a smartphone app and applies spatio-temporal clustering techniques and cross-correlation analysis to create maps of abnormal symptomatology incidence that are made publicly available. The results of the cross-correlation analysis identify optimal temporal lags between symptoms and a range of COVID-19 outcomes, with new taste/smell loss showing the highest correlations. We also identified temporal clusters of change in taste/smell entries and confirmed COVID-19 incidence in Baltimore City and County. Further, we utilized an extended simulated dataset to showcase our analytics in Maryland. The resulting clusters can serve as indicators of emerging COVID-19 outbreaks, and support syndromic surveillance as an early warning system for disease prevention and control.

Abstract Background There is uncertainty regarding the effect of anemia and red blood cell transfusion on functional outcome following acute ischemic stroke. We studied the relationship of hemoglobin parameters and red cell transfusion with post stroke functional outcome after adjustment for neurological severity and medical comorbidities. Methods Retrospective cohort study of 536 patients discharged with a diagnosis of ischemic stroke from a tertiary care hospital between January 2012 and April 2015. Hemoglobin level at hospital admission, lowest recorded value during hospitalization (nadir), delta hemoglobin (admission minus nadir), red cell transfusion during hospitalization were noted. Charlson Comorbidity Index (CCI) was computed as a summary measure of medical comorbidities. A multivariable logistic regression model was used to determine risk-adjusted odds of unfavorable outcome, defined as a modified Rankin Score of > 2. Results Anemia was present on hospital admission in 31% of patients. Forty five percent of patients had unfavorable outcome. In the univariable analysis increasing age, admission National Institutes of Health Stroke Scale (NIHSS), CCI, nadir hemoglobin, delta hemoglobin and blood transfusion were associated with unfavorable outcome. In the multivariable model, only increasing age, CCI and NIHSS remained associated with unfavorable outcome. No quadratic association was found on repeating the model to identify a possible U-shaped relationship of hemoglobin with outcome. Conclusions Our findings contradict prior observational studies and highlight an area of clinical equipoise regarding the optimal management of anemia in patients hospitalized for ischemic stroke. This uncertainty could be addressed with appropriately designed clinical trials.

Summary: Acyl CoA metabolites derived from the catabolism of carbon fuels can react with lysine residues of mitochondrial proteins, giving rise to a large family of post-translational modifications (PTMs). Mass spectrometry-based detection of thousands of acyl-PTMs scattered throughout the proteome has established a strong link between mitochondrial hyperacylation and cardiometabolic diseases; however, the functional consequences of these modifications remain uncertain. Here, we use a comprehensive respiratory diagnostics platform to evaluate three disparate models of mitochondrial hyperacylation in the mouse heart caused by genetic deletion of malonyl CoA decarboxylase (MCD), SIRT5 demalonylase and desuccinylase, or SIRT3 deacetylase. In each case, elevated acylation is accompanied by marginal respiratory phenotypes. Of the >60 mitochondrial energy fluxes evaluated, the only outcome consistently observed across models is a ∼15% decrease in ATP synthase activity. In sum, the findings suggest that the vast majority of mitochondrial acyl PTMs occur as stochastic events that minimally affect mitochondrial bioenergetics. : Fisher-Wellman et al. use a recently developed mitochondrial diagnostics platform for deep phenotyping of heart mitochondria derived from three disparate genetic models of protein hyperacylation. Their findings oppose the notion that hyperacylation of the mitochondrial proteome leads to broad-ranging vulnerabilities in respiratory function and bioenergetics. Keywords: mitochondrial diagnostics, lysine acylation, malonylation, ATP synthase

Objective: A branched-chain amino acid (BCAA)-related metabolic signature is strongly associated with insulin resistance and predictive of incident diabetes and intervention outcomes. To better understand the role that this metabolite cluster plays in obesity-related metabolic dysfunction, we studied the impact of BCAA restriction in a rodent model of obesity in which BCAA metabolism is perturbed in ways that mirror the human condition. Methods: Zucker-lean rats (ZLR) and Zucker-fatty rats (ZFR) were fed either a custom control, low fat (LF) diet, or an isonitrogenous, isocaloric LF diet in which all three BCAA (Leu, Ile, Val) were reduced by 45% (LF-RES). We performed comprehensive metabolic and physiologic profiling to characterize the effects of BCAA restriction on energy balance, insulin sensitivity, and glucose, lipid and amino acid metabolism. Results: LF-fed ZFR had higher levels of circulating BCAA and lower levels of glycine compared to LF-fed ZLR. Feeding ZFR with the LF-RES diet lowered circulating BCAA to levels found in LF-fed ZLR. Activity of the rate limiting enzyme in the BCAA catabolic pathway, branched chain keto acid dehydrogenase (BCKDH), was lower in liver but higher in skeletal muscle of ZFR compared to ZLR and was not responsive to diet in either tissue. BCAA restriction had very little impact on metabolites studied in liver of ZFR where BCAA content was low, and BCKDH activity was suppressed. However, in skeletal muscle of LF-fed ZFR compared to LF-fed ZLR, where BCAA content and BCKDH activity were increased, accumulation of fatty acyl CoAs was completely normalized by dietary BCAA restriction. BCAA restriction also normalized skeletal muscle glycine content and increased urinary acetyl glycine excretion in ZFR. These effects were accompanied by lower RER and improved skeletal muscle insulin sensitivity in LF-RES fed ZFR as measured by hyperinsulinemic-isoglycemic clamp. Conclusions: Our data are consistent with a model wherein elevated circulating BCAA contribute to development of obesity-related insulin resistance by interfering with lipid oxidation in skeletal muscle. BCAA-dependent lowering of the skeletal muscle glycine pool appears to contribute to this effect by slowing acyl-glycine export to the urine. Keywords: Obesity, BCAA, Insulin sensitivity, Metabolism

Lysine acetylation (AcK), a posttranslational modification wherein a two-carbon acetyl group binds covalently to a lysine residue, occurs prominently on mitochondrial proteins and has been linked to metabolic dysfunction. An emergent theory suggests mitochondrial AcK occurs via mass action rather than targeted catalysis. To test this hypothesis, we performed mass spectrometry-based acetylproteomic analyses of quadriceps muscles from mice with skeletal muscle-specific deficiency of carnitine acetyltransferase (CrAT), an enzyme that buffers the mitochondrial acetyl-CoA pool by converting short-chain acyl-CoAs to their membrane permeant acylcarnitine counterparts. CrAT deficiency increased tissue acetyl-CoA levels and susceptibility to diet-induced AcK of broad-ranging mitochondrial proteins, coincident with diminished whole body glucose control. Sub-compartment acetylproteome analyses of muscles from obese mice and humans showed remarkable overrepresentation of mitochondrial matrix proteins. These findings reveal roles for CrAT and L-carnitine in modulating the muscle acetylproteome and provide strong experimental evidence favoring the nonenzymatic carbon pressure model of mitochondrial AcK.

Palmitic acid (PA) is associated with higher blood concentrations of medium-chain acylcarnitines (MCACs), and we hypothesized that PA may inhibit progression of FA β-oxidation. Using a cross-over design, 17 adults were fed high PA (HPA) and low PA/high oleic acid (HOA) diets, each for 3 weeks. The [1-13C]PA and [13-13C]PA tracers were administered with food in random order with each diet, and we assessed PA oxidation (PA OX) and serum AC concentration to determine whether a higher PA intake promoted incomplete PA OX. Dietary PA was completely oxidized during the HOA diet, but only about 40% was oxidized during the HPA diet. The [13-13C]PA/[1-13C]PA ratio of PA OX had an approximate value of 1.0 for either diet, but the ratio of the serum concentrations of MCACs to long-chain ACs (LCACs) was significantly higher during the HPA diet. Thus, direct measurement of PA OX did not confirm that the HPA diet caused incomplete PA OX, despite the modest, but statistically significant, increase in the ratio of MCACs to LCACs in blood.

Nonalcoholic fatty liver disease is a major public health concern in the obese and type 2 diabetic populations. The high-fat lard diet induces obesity and fatty liver in C57BL/6J mice and suppresses expression of the PPAR-target gene, FA elongase 5 (Elovl5). Elovl5 plays a key role in MUFA and PUFA synthesis. Increasing hepatic Elovl5 activity in obese mice lowered hepatic TGs and endoplasmic reticulum stress markers (X-box binding protein 1 and cAMP-dependent transcription factor 6α) and increased TG catabolism and fatty acyl carnitines. Increased hepatic Elovl5 activity did not increase hepatic capacity for β-oxidation. Elovl5 effects on hepatic TG catabolism were linked to increased protein levels of adipocyte TG lipase (ATGL) and comparative gene identification 58 (CGI58). Elevated hepatic Elovl5 activity also induced the expression of some (pyruvate dehydrogenase kinase 4 and fibroblast growth factor 21), but not other cytochrome P450 4A10 (CYP4A10), PPAR-target genes. FA products of Elovl5 activity increased ATGL, but not CGI58, mRNA through PPARβ-dependent mechanisms in human HepG2 cells. Treatment of mouse AML12 hepatocytes with the PPARβ agonist (GW0742) decreased 14C-18:2,n-6 in TGs but did not affect β-oxidation. These studies establish that Elovl5 activity regulates hepatic levels of FAs controlling PPARβ activity, ATGL expression, and TG catabolism, but not FA oxidation.

BackgroundMetabolic impairment is an important contributor to heart failure (HF) pathogenesis and progression. Dysregulated metabolic pathways remain poorly characterized in patients with HF and preserved ejection fraction (HFpEF). We sought to determine metabolic abnormalities in HFpEF and identify pathways differentially altered in HFpEF versus HF with reduced ejection fraction (HFrEF). Methods and ResultsWe identified HFpEF cases, HFrEF controls, and no‐HF controls from the CATHGEN study of sequential patients undergoing cardiac catheterization. HFpEF cases (N=282) were defined by left ventricular ejection fraction (LVEF) ≥45%, diastolic dysfunction grade ≥1, and history of HF; HFrEF controls (N=279) were defined similarly, except for having LVEF

Mitochondrial dysfunction has been implicated in the pathogenesis of type 2 diabetes. Identifying novel regulators of mitochondrial bioenergetics will broaden our understanding of regulatory checkpoints that coordinate complex metabolic pathways. We previously showed that Nur77, an orphan nuclear receptor of the NR4A family, regulates the expression of genes linked to glucose utilization. Here we demonstrate that expression of Nur77 in skeletal muscle also enhances mitochondrial function. We generated MCK-Nur77 transgenic mice that express wild-type Nur77 specifically in skeletal muscle. Nur77-overexpressing muscle had increased abundance of oxidative muscle fibers and mitochondrial DNA content. Transgenic muscle also exhibited enhanced oxidative metabolism, suggestive of increased mitochondrial activity. Metabolomic analysis confirmed that Nur77 transgenic muscle favored fatty acid oxidation over glucose oxidation, mimicking the metabolic profile of fasting. Nur77 expression also improved the intrinsic respiratory capacity of isolated mitochondria, likely due to the increased abundance of complex I of the electron transport chain. These changes in mitochondrial metabolism translated to improved muscle contractile function ex vivo and improved cold tolerance in vivo. Our studies outline a novel role for Nur77 in the regulation of oxidative metabolism and mitochondrial activity in skeletal muscle.

We previously reported how the loss of CHIP expression (Carboxyl terminus of Hsc70-Interacting Protein) during pressure overload resulted in robust cardiac dysfunction, which was accompanied by a failure to maintain ATP levels in the face of increased energy demand. In this study, we analyzed the cardiac metabolome after seven days of pressure overload and found an increase in long-chain and medium-chain fatty acid metabolites in wild-type hearts. This response was attenuated in mice that lack expression of CHIP (CHIP−/−). These findings suggest that CHIP may play an essential role in regulating oxidative metabolism pathways that are regulated, in part, by the nuclear receptor PPARα (Peroxisome Proliferator-Activated Receptor alpha). Next, we challenged CHIP−/− mice with the PPARα agonist called fenofibrate. We found that treating CHIP−/− mice with fenofibrate for five weeks under non-pressure overload conditions resulted in decreased skeletal muscle mass, compared to wild-type mice, and a marked increase in cardiac fibrosis accompanied by a decrease in cardiac function. Fenofibrate resulted in decreased mitochondrial cristae density in CHIP−/− hearts as well as decreased expression of genes involved in the initiation of autophagy and mitophagy, which suggests that a metabolic challenge, in the absence of CHIP expression, impacts pathways that contribute to mitochondrial quality control. In conclusion, in the absence of functional CHIP expression, fenofibrate results in unexpected skeletal muscle and cardiac pathologies. These findings are particularly relevant to patients harboring loss-of-function mutations in CHIP and are consistent with a prominent role for CHIP in regulating cardiac metabolism.